We are requesting funds to analyze data recently collected for a nested case-control study of prostate cancer incidence and exposure to metal working fluids (MWF). This proposed research will follow up preliminary data suggesting an association between prostate cancer mortality and MWF in a cohort study of automobile manufacturing workers with extensive quantitative information on exposure to straight, soluble and synthetic MWF. We have identified incident cases in the cohort through the state cancer registries and are proposing to analyze the case-control data in order to examine the hypothesis that oil or chemical components of these fluids may interfere with hormonal activity of the prostate, disrupt its normal function and contribute to the development of prostate cancer. We propose to use both parametric and nonparametric statistical models to examine the exposure-response relationships. Smoothing techniques based on penalized splines will be used in the nonparametric models to relax the assumption of linearity between exposure and risk. Exposure will be based on two different sets of exposure time windows. Exposure will be cumulated in consecutive windows defined by (1) periods of hormonal changes during a man's life (puberty to 35, 35-50 and > 50 years of age), and (2) suspected latency periods for prostate cancer, (10, 20, and 30 years prior to age at diagnosis). We will explore the healthy worker survivor effect by comparing three methods for reducing the bias, lagging cumulative exposure for 5 years to exclude recent exposures incurred by those who remained on the job, adjusting for employment status at the age of diagnosis, and controlling for time since hire as potential confounder. Because the cohort is 16% African-American, we will also examine race as a risk factor in prostate cancer incidence and survival.